ai-document-parser 0.1.101 → 0.1.104
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +3 -2
- package/dist/convert/blocks.d.ts +14 -0
- package/dist/convert/blocks.d.ts.map +1 -0
- package/dist/convert/blocks.js +254 -0
- package/dist/convert/blocks.js.map +1 -0
- package/dist/convert/converter.d.ts.map +1 -1
- package/dist/convert/converter.js +84 -62
- package/dist/convert/converter.js.map +1 -1
- package/dist/convert/emit.d.ts +1 -0
- package/dist/convert/emit.d.ts.map +1 -1
- package/dist/convert/emit.js +32 -33
- package/dist/convert/emit.js.map +1 -1
- package/dist/edit/docx-ooxml.d.ts.map +1 -1
- package/dist/edit/docx-ooxml.js +5 -33
- package/dist/edit/docx-ooxml.js.map +1 -1
- package/dist/index.d.ts +1 -1
- package/dist/index.d.ts.map +1 -1
- package/dist/parse/docx-blocks.d.ts +18 -0
- package/dist/parse/docx-blocks.d.ts.map +1 -0
- package/dist/parse/docx-blocks.js +654 -0
- package/dist/parse/docx-blocks.js.map +1 -0
- package/dist/parse/html-blocks.d.ts +14 -0
- package/dist/parse/html-blocks.d.ts.map +1 -0
- package/dist/parse/html-blocks.js +244 -0
- package/dist/parse/html-blocks.js.map +1 -0
- package/dist/parse/pptx.d.ts +0 -1
- package/dist/parse/pptx.d.ts.map +1 -1
- package/dist/parse/pptx.js +95 -24
- package/dist/parse/pptx.js.map +1 -1
- package/dist/parse/router.d.ts.map +1 -1
- package/dist/parse/router.js +25 -6
- package/dist/parse/router.js.map +1 -1
- package/dist/types.d.ts +37 -0
- package/dist/types.d.ts.map +1 -1
- package/package.json +2 -2
package/dist/convert/emit.js
CHANGED
|
@@ -1,6 +1,8 @@
|
|
|
1
1
|
import { csvEscapeCell, escapeHtml } from '../utils/text.js';
|
|
2
2
|
import { spreadsheetToText } from '../edit/spreadsheet.js';
|
|
3
|
+
import { blocksFromSpreadsheet, emitHtmlFromBlocks, emitMarkdownFromBlocks, } from './blocks.js';
|
|
3
4
|
export { csvEscapeCell, escapeHtml } from '../utils/text.js';
|
|
5
|
+
export { emitMarkdownFromBlocks, emitHtmlFromBlocks, emitMarkdownTable, blocksFromSpreadsheet, blocksFromCsvRows, flattenBlocksToParagraphs, } from './blocks.js';
|
|
4
6
|
/** Emit markdown / html / text / csv from already-extracted content. */
|
|
5
7
|
export function emitText(text) {
|
|
6
8
|
return text.replace(/\r\n/g, '\n').trim();
|
|
@@ -37,43 +39,40 @@ export function emitCsvFromSheet(structured, sheetName) {
|
|
|
37
39
|
.join('\n');
|
|
38
40
|
}
|
|
39
41
|
export function emitMarkdownFromSpreadsheet(structured) {
|
|
40
|
-
|
|
41
|
-
for (const [name, rows] of Object.entries(structured.sheets)) {
|
|
42
|
-
parts.push(`## ${name}`);
|
|
43
|
-
if (!rows.length)
|
|
44
|
-
continue;
|
|
45
|
-
const header = rows[0].map((c) => String(c ?? ''));
|
|
46
|
-
const sep = header.map(() => '---');
|
|
47
|
-
parts.push(`| ${header.join(' | ')} |`);
|
|
48
|
-
parts.push(`| ${sep.join(' | ')} |`);
|
|
49
|
-
for (const row of rows.slice(1)) {
|
|
50
|
-
parts.push(`| ${row.map((c) => String(c ?? '')).join(' | ')} |`);
|
|
51
|
-
}
|
|
52
|
-
parts.push('');
|
|
53
|
-
}
|
|
54
|
-
return parts.join('\n').trim();
|
|
42
|
+
return emitMarkdownFromBlocks(blocksFromSpreadsheet(structured.sheets));
|
|
55
43
|
}
|
|
56
44
|
export function emitMarkdownFromPptx(slides) {
|
|
57
|
-
|
|
58
|
-
|
|
59
|
-
|
|
60
|
-
|
|
61
|
-
|
|
62
|
-
|
|
63
|
-
|
|
45
|
+
const blocks = [];
|
|
46
|
+
slides.forEach((texts, i) => {
|
|
47
|
+
blocks.push({
|
|
48
|
+
type: 'heading',
|
|
49
|
+
level: 2,
|
|
50
|
+
runs: [{ text: `Slide ${i + 1}` }],
|
|
51
|
+
});
|
|
52
|
+
texts.forEach((t, j) => {
|
|
53
|
+
if (j === 0) {
|
|
54
|
+
blocks.push({ type: 'heading', level: 3, runs: [{ text: t }] });
|
|
55
|
+
}
|
|
56
|
+
else {
|
|
57
|
+
blocks.push({ type: 'paragraph', runs: [{ text: t }] });
|
|
58
|
+
}
|
|
59
|
+
});
|
|
60
|
+
});
|
|
61
|
+
return emitMarkdownFromBlocks(blocks);
|
|
64
62
|
}
|
|
65
63
|
export function emitHtmlFromPptx(slides) {
|
|
66
|
-
const
|
|
67
|
-
|
|
68
|
-
|
|
69
|
-
|
|
70
|
-
|
|
71
|
-
|
|
72
|
-
|
|
73
|
-
|
|
74
|
-
|
|
75
|
-
|
|
76
|
-
|
|
64
|
+
const blocks = [];
|
|
65
|
+
slides.forEach((texts, i) => {
|
|
66
|
+
blocks.push({
|
|
67
|
+
type: 'heading',
|
|
68
|
+
level: 2,
|
|
69
|
+
runs: [{ text: `Slide ${i + 1}` }],
|
|
70
|
+
});
|
|
71
|
+
for (const t of texts) {
|
|
72
|
+
blocks.push({ type: 'paragraph', runs: [{ text: t }] });
|
|
73
|
+
}
|
|
74
|
+
});
|
|
75
|
+
return emitHtmlFromBlocks(blocks);
|
|
77
76
|
}
|
|
78
77
|
export function emitTextFromSpreadsheet(structured) {
|
|
79
78
|
return spreadsheetToText(structured);
|
package/dist/convert/emit.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"emit.js","sourceRoot":"","sources":["../../src/convert/emit.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAE7D,OAAO,EAAE,iBAAiB,EAAE,MAAM,wBAAwB,CAAC;
|
|
1
|
+
{"version":3,"file":"emit.js","sourceRoot":"","sources":["../../src/convert/emit.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAE7D,OAAO,EAAE,iBAAiB,EAAE,MAAM,wBAAwB,CAAC;AAC3D,OAAO,EACL,qBAAqB,EACrB,kBAAkB,EAClB,sBAAsB,GAEvB,MAAM,aAAa,CAAC;AAErB,OAAO,EAAE,aAAa,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAC7D,OAAO,EACL,sBAAsB,EACtB,kBAAkB,EAClB,iBAAiB,EACjB,qBAAqB,EACrB,iBAAiB,EACjB,yBAAyB,GAC1B,MAAM,aAAa,CAAC;AAErB,wEAAwE;AACxE,MAAM,UAAU,QAAQ,CAAC,IAAY;IACnC,OAAO,IAAI,CAAC,OAAO,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;AAC5C,CAAC;AAED,MAAM,UAAU,0BAA0B,CAAC,UAAoB;IAC7D,OAAO,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;AACtE,CAAC;AAED,MAAM,UAAU,sBAAsB,CAAC,UAAoB;IACzD,MAAM,IAAI,GAAG,UAAU;SACpB,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;SACpB,MAAM,CAAC,OAAO,CAAC;SACf,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,MAAM,UAAU,CAAC,CAAC,CAAC,MAAM,CAAC;SACrC,IAAI,CAAC,IAAI,CAAC,CAAC;IACd,OAAO,kCAAkC,IAAI,oBAAoB,CAAC;AACpE,CAAC;AAED,MAAM,UAAU,oBAAoB,CAAC,IAAY;IAC/C,OAAO,QAAQ,CAAC,IAAI,CAAC,CAAC;AACxB,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,IAAY;IAC3C,MAAM,UAAU,GAAG,IAAI;SACpB,KAAK,CAAC,SAAS,CAAC;SAChB,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;SACpB,MAAM,CAAC,OAAO,CAAC,CAAC;IACnB,IAAI,CAAC,UAAU,CAAC,MAAM,EAAE,CAAC;QACvB,OAAO,qCAAqC,UAAU,CAAC,IAAI,CAAC,wBAAwB,CAAC;IACvF,CAAC;IACD,OAAO,sBAAsB,CAAC,UAAU,CAAC,CAAC;AAC5C,CAAC;AAED,MAAM,UAAU,gBAAgB,CAC9B,UAAiC,EACjC,SAAkB;IAElB,MAAM,IAAI,GAAG,SAAS,IAAI,MAAM,CAAC,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC;IAC5D,MAAM,IAAI,GAAG,UAAU,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC;IAC3C,OAAO,IAAI;SACR,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;SACxD,IAAI,CAAC,IAAI,CAAC,CAAC;AAChB,CAAC;AAED,MAAM,UAAU,2BAA2B,CAAC,UAAiC;IAC3E,OAAO,sBAAsB,CAAC,qBAAqB,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC;AAC1E,CAAC;AAED,MAAM,UAAU,oBAAoB,CAAC,MAAkB;IACrD,MAAM,MAAM,GAAmB,EAAE,CAAC;IAClC,MAAM,CAAC,OAAO,CAAC,CAAC,KAAK,EAAE,CAAC,EAAE,EAAE;QAC1B,MAAM,CAAC,IAAI,CAAC;YACV,IAAI,EAAE,SAAS;YACf,KAAK,EAAE,CAAC;YACR,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,SAAS,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC;SACnC,CAAC,CAAC;QACH,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;YACrB,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;gBACZ,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;YAClE,CAAC;iBAAM,CAAC;gBACN,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;YAC1D,CAAC;QACH,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IACH,OAAO,sBAAsB,CAAC,MAAM,CAAC,CAAC;AACxC,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,MAAkB;IACjD,MAAM,MAAM,GAAmB,EAAE,CAAC;IAClC,MAAM,CAAC,OAAO,CAAC,CAAC,KAAK,EAAE,CAAC,EAAE,EAAE;QAC1B,MAAM,CAAC,IAAI,CAAC;YACV,IAAI,EAAE,SAAS;YACf,KAAK,EAAE,CAAC;YACR,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,SAAS,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC;SACnC,CAAC,CAAC;QACH,KAAK,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC;YACtB,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;QAC1D,CAAC;IACH,CAAC,CAAC,CAAC;IACH,OAAO,kBAAkB,CAAC,MAAM,CAAC,CAAC;AACpC,CAAC;AAED,MAAM,UAAU,uBAAuB,CAAC,UAAiC;IACvE,OAAO,iBAAiB,CAAC,UAAU,CAAC,CAAC;AACvC,CAAC;AAED,MAAM,UAAU,SAAS,CAAC,IAAY,EAAE,QAAQ,GAAG,IAAI;IACrD,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC;IAC5B,IAAI,CAAC,OAAO;QAAE,OAAO,EAAE,CAAC;IACxB,IAAI,OAAO,CAAC,MAAM,IAAI,QAAQ;QAAE,OAAO,CAAC,OAAO,CAAC,CAAC;IACjD,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,OAAO,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC;QAC1B,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,CAAC,CAAC;QAC5C,CAAC,IAAI,QAAQ,CAAC;IAChB,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"docx-ooxml.d.ts","sourceRoot":"","sources":["../../src/edit/docx-ooxml.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;
|
|
1
|
+
{"version":3,"file":"docx-ooxml.d.ts","sourceRoot":"","sources":["../../src/edit/docx-ooxml.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AASlD,wBAAgB,SAAS,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAO9C;AAkGD,wBAAsB,QAAQ,CAC5B,QAAQ,EAAE,MAAM,EAChB,SAAS,EAAE,MAAM,EACjB,SAAS,EAAE,MAAM,EACjB,UAAU,EAAE,OAAO,GAClB,OAAO,CAAC;IAAE,OAAO,EAAE,MAAM,CAAC;IAAC,OAAO,EAAE,MAAM,CAAA;CAAE,CAAC,CAyC/C;AAED,wBAAsB,oBAAoB,CACxC,QAAQ,EAAE,MAAM,EAChB,UAAU,EAAE,MAAM,EAAE,GACnB,OAAO,CAAC;IAAE,OAAO,EAAE,MAAM,CAAC;IAAC,OAAO,EAAE,MAAM,CAAA;CAAE,CAAC,CA4B/C;AAED,wBAAsB,kBAAkB,CACtC,QAAQ,EAAE,MAAM,GACf,OAAO,CAAC,cAAc,CAAC,CAMzB;AAED,wBAAsB,4BAA4B,CAChD,GAAG,EAAE,MAAM,GACV,OAAO,CAAC,cAAc,CAAC,CAMzB"}
|
package/dist/edit/docx-ooxml.js
CHANGED
|
@@ -1,6 +1,7 @@
|
|
|
1
1
|
import { readFile } from 'node:fs/promises';
|
|
2
2
|
import JSZip from 'jszip';
|
|
3
3
|
import { DocumentError } from '../utils/errors.js';
|
|
4
|
+
import { docxStructuredFromZip } from '../parse/docx-blocks.js';
|
|
4
5
|
const DOC_PARTS = [
|
|
5
6
|
'word/document.xml',
|
|
6
7
|
/^word\/header\d*\.xml$/,
|
|
@@ -14,31 +15,6 @@ export function escapeXml(text) {
|
|
|
14
15
|
.replace(/"/g, '"')
|
|
15
16
|
.replace(/'/g, ''');
|
|
16
17
|
}
|
|
17
|
-
function decodeXmlEntities(s) {
|
|
18
|
-
return s
|
|
19
|
-
.replace(/</g, '<')
|
|
20
|
-
.replace(/>/g, '>')
|
|
21
|
-
.replace(/"/g, '"')
|
|
22
|
-
.replace(/'/g, "'")
|
|
23
|
-
.replace(/&/g, '&');
|
|
24
|
-
}
|
|
25
|
-
function paragraphTextsFromXml(xml) {
|
|
26
|
-
const paragraphs = [];
|
|
27
|
-
const paraRegex = /<w:p[\s\S]*?<\/w:p>/g;
|
|
28
|
-
let pm;
|
|
29
|
-
while ((pm = paraRegex.exec(xml)) !== null) {
|
|
30
|
-
const wtRegex = /<w:t(?:\s[^>]*)?>([^<]*)<\/w:t>/g;
|
|
31
|
-
const parts = [];
|
|
32
|
-
let tm;
|
|
33
|
-
while ((tm = wtRegex.exec(pm[0])) !== null) {
|
|
34
|
-
parts.push(decodeXmlEntities(tm[1]));
|
|
35
|
-
}
|
|
36
|
-
const text = parts.join('').trim();
|
|
37
|
-
if (text)
|
|
38
|
-
paragraphs.push(text);
|
|
39
|
-
}
|
|
40
|
-
return paragraphs;
|
|
41
|
-
}
|
|
42
18
|
function isDocPart(name) {
|
|
43
19
|
return DOC_PARTS.some((p) => typeof p === 'string' ? name === p : p.test(name));
|
|
44
20
|
}
|
|
@@ -163,20 +139,16 @@ export async function appendDocxParagraphs(filePath, paragraphs) {
|
|
|
163
139
|
}
|
|
164
140
|
export async function readDocxStructured(filePath) {
|
|
165
141
|
const zip = await loadZip(filePath);
|
|
166
|
-
|
|
167
|
-
if (!docFile) {
|
|
142
|
+
if (!zip.file('word/document.xml')) {
|
|
168
143
|
throw new DocumentError('PARSE_FAILED', 'Invalid docx: missing word/document.xml');
|
|
169
144
|
}
|
|
170
|
-
|
|
171
|
-
return { format: 'docx', paragraphs: paragraphTextsFromXml(xml) };
|
|
145
|
+
return docxStructuredFromZip(zip);
|
|
172
146
|
}
|
|
173
147
|
export async function readDocxStructuredFromBuffer(buf) {
|
|
174
148
|
const zip = await JSZip.loadAsync(buf);
|
|
175
|
-
|
|
176
|
-
if (!docFile) {
|
|
149
|
+
if (!zip.file('word/document.xml')) {
|
|
177
150
|
throw new DocumentError('PARSE_FAILED', 'Invalid docx: missing word/document.xml');
|
|
178
151
|
}
|
|
179
|
-
|
|
180
|
-
return { format: 'docx', paragraphs: paragraphTextsFromXml(xml) };
|
|
152
|
+
return docxStructuredFromZip(zip);
|
|
181
153
|
}
|
|
182
154
|
//# sourceMappingURL=docx-ooxml.js.map
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"docx-ooxml.js","sourceRoot":"","sources":["../../src/edit/docx-ooxml.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,EAAE,aAAa,EAAE,MAAM,oBAAoB,CAAC;
|
|
1
|
+
{"version":3,"file":"docx-ooxml.js","sourceRoot":"","sources":["../../src/edit/docx-ooxml.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,EAAE,aAAa,EAAE,MAAM,oBAAoB,CAAC;AAEnD,OAAO,EAAE,qBAAqB,EAAE,MAAM,yBAAyB,CAAC;AAEhE,MAAM,SAAS,GAA2B;IACxC,mBAAmB;IACnB,wBAAwB;IACxB,wBAAwB;CACzB,CAAC;AAEF,MAAM,UAAU,SAAS,CAAC,IAAY;IACpC,OAAO,IAAI;SACR,OAAO,CAAC,IAAI,EAAE,OAAO,CAAC;SACtB,OAAO,CAAC,IAAI,EAAE,MAAM,CAAC;SACrB,OAAO,CAAC,IAAI,EAAE,MAAM,CAAC;SACrB,OAAO,CAAC,IAAI,EAAE,QAAQ,CAAC;SACvB,OAAO,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAC;AAC7B,CAAC;AAED,SAAS,SAAS,CAAC,IAAY;IAC7B,OAAO,SAAS,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAC1B,OAAO,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAClD,CAAC;AACJ,CAAC;AAED,KAAK,UAAU,OAAO,CAAC,QAAgB;IACrC,MAAM,GAAG,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACrC,OAAO,KAAK,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC;AAC9B,CAAC;AAED,SAAS,qBAAqB,CAC5B,OAAe,EACf,SAAiB,EACjB,SAAiB,EACjB,UAAmB;IAEnB,MAAM,OAAO,GAAG,kCAAkC,CAAC;IACnD,MAAM,IAAI,GAAqC,EAAE,CAAC;IAClD,IAAI,CAAyB,CAAC;IAC9B,OAAO,CAAC,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC5C,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;IACxC,CAAC;IACD,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACtB,OAAO,EAAE,GAAG,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;IACpC,CAAC;IAED,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAClD,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,SAAS,CAAC,EAAE,CAAC;QAClC,OAAO,EAAE,GAAG,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;IACpC,CAAC;IAED,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,eAAe,GAAG,QAAQ,CAAC;IAC/B,IAAI,UAAU,EAAE,CAAC;QACf,MAAM,KAAK,GAAG,QAAQ,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC;QACxC,KAAK,GAAG,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC;QACzB,eAAe,GAAG,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAC1C,CAAC;SAAM,CAAC;QACN,MAAM,GAAG,GAAG,QAAQ,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;QACxC,IAAI,GAAG,KAAK,CAAC,CAAC;YAAE,OAAO,EAAE,GAAG,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;QAClD,KAAK,GAAG,CAAC,CAAC;QACV,eAAe;YACb,QAAQ,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC;gBACtB,SAAS;gBACT,QAAQ,CAAC,KAAK,CAAC,GAAG,GAAG,SAAS,CAAC,MAAM,CAAC,CAAC;IAC3C,CAAC;IAED,MAAM,OAAO,GAAG,SAAS,CAAC,eAAe,CAAC,CAAC;IAC3C,MAAM,MAAM,GAAG,kCAAkC,OAAO,cAAc,CAAC;IACvE,MAAM,QAAQ,GAAG,OAAO,CAAC,KAAK,CAAC,yBAAyB,CAAC,CAAC;IAC1D,MAAM,GAAG,GAAG,QAAQ,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;IACxC,OAAO,EAAE,GAAG,EAAE,QAAQ,GAAG,GAAG,MAAM,QAAQ,EAAE,KAAK,EAAE,CAAC;AACtD,CAAC;AAED,SAAS,oBAAoB,CAC3B,GAAW,EACX,SAAiB,EACjB,SAAiB,EACjB,UAAmB;IAEnB,MAAM,SAAS,GAAG,sBAAsB,CAAC;IACzC,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,MAAM,MAAM,GAAG,GAAG,CAAC,OAAO,CAAC,SAAS,EAAE,CAAC,IAAI,EAAE,EAAE;QAC7C,MAAM,EAAE,GAAG,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,qBAAqB,CAChD,IAAI,EACJ,SAAS,EACT,SAAS,EACT,UAAU,CACX,CAAC;QACF,KAAK,IAAI,KAAK,CAAC;QACf,OAAO,IAAI,CAAC;IACd,CAAC,CAAC,CAAC;IAEH,IAAI,KAAK,KAAK,CAAC,IAAI,GAAG,CAAC,QAAQ,CAAC,SAAS,CAAC,EAAE,CAAC;QAC3C,IAAI,UAAU,EAAE,CAAC;YACf,MAAM,KAAK,GAAG,GAAG,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC;YACnC,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBACrB,OAAO,EAAE,GAAG,EAAE,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC,EAAE,KAAK,EAAE,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAC5E,CAAC;QACH,CAAC;QACD,MAAM,GAAG,GAAG,GAAG,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;QACnC,IAAI,GAAG,KAAK,CAAC,CAAC,EAAE,CAAC;YACf,OAAO;gBACL,GAAG,EACD,GAAG,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC;oBACjB,SAAS,CAAC,SAAS,CAAC;oBACpB,GAAG,CAAC,KAAK,CAAC,GAAG,GAAG,SAAS,CAAC,MAAM,CAAC;gBACnC,KAAK,EAAE,CAAC;aACT,CAAC;QACJ,CAAC;IACH,CAAC;IAED,OAAO,EAAE,GAAG,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,CAAC;AACvC,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,QAAQ,CAC5B,QAAgB,EAChB,SAAiB,EACjB,SAAiB,EACjB,UAAmB;IAEnB,IAAI,CAAC,SAAS,EAAE,CAAC;QACf,MAAM,IAAI,aAAa,CAAC,kBAAkB,EAAE,2BAA2B,CAAC,CAAC;IAC3E,CAAC;IAED,MAAM,GAAG,GAAG,MAAM,OAAO,CAAC,QAAQ,CAAC,CAAC;IACpC,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,KAAK,MAAM,IAAI,IAAI,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC;QAC1C,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC,IAAI,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,GAAG;YAAE,SAAS;QACtD,MAAM,GAAG,GAAG,MAAM,GAAG,CAAC,IAAI,CAAC,IAAI,CAAE,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;QAClD,MAAM,EAAE,GAAG,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,oBAAoB,CAC/C,GAAG,EACH,SAAS,EACT,SAAS,EACT,UAAU,CACX,CAAC;QACF,KAAK,IAAI,KAAK,CAAC;QACf,GAAG,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;IACvB,CAAC;IAED,IAAI,KAAK,KAAK,CAAC,EAAE,CAAC;QAChB,MAAM,IAAI,aAAa,CACrB,uBAAuB,EACvB,8EAA8E,CAC/E,CAAC;IACJ,CAAC;IACD,IAAI,KAAK,GAAG,CAAC,IAAI,CAAC,UAAU,EAAE,CAAC;QAC7B,MAAM,IAAI,aAAa,CACrB,kBAAkB,EAClB,SAAS,KAAK,qDAAqD,CACpE,CAAC;IACJ,CAAC;IAED,MAAM,EAAE,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC,kBAAkB,CAAC,CAAC;IACvD,MAAM,GAAG,GAAG,MAAM,GAAG,CAAC,aAAa,CAAC,EAAE,IAAI,EAAE,YAAY,EAAE,CAAC,CAAC;IAC5D,MAAM,SAAS,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;IAC/B,OAAO;QACL,OAAO,EAAE,iBAAiB,QAAQ,KAAK,KAAK,kDAAkD;QAC9F,OAAO,EAAE,KAAK;KACf,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,oBAAoB,CACxC,QAAgB,EAChB,UAAoB;IAEpB,MAAM,GAAG,GAAG,MAAM,OAAO,CAAC,QAAQ,CAAC,CAAC;IACpC,MAAM,OAAO,GAAG,GAAG,CAAC,IAAI,CAAC,mBAAmB,CAAC,CAAC;IAC9C,IAAI,CAAC,OAAO,EAAE,CAAC;QACb,MAAM,IAAI,aAAa,CAAC,cAAc,EAAE,yCAAyC,CAAC,CAAC;IACrF,CAAC;IAED,IAAI,GAAG,GAAG,MAAM,OAAO,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;IACxC,MAAM,MAAM,GAAG,UAAU;SACtB,GAAG,CACF,CAAC,CAAC,EAAE,EAAE,CACJ,uCAAuC,SAAS,CAAC,CAAC,IAAI,GAAG,CAAC,oBAAoB,CACjF;SACA,IAAI,CAAC,EAAE,CAAC,CAAC;IAEZ,IAAI,CAAC,GAAG,CAAC,QAAQ,CAAC,WAAW,CAAC,EAAE,CAAC;QAC/B,MAAM,IAAI,aAAa,CAAC,cAAc,EAAE,uCAAuC,CAAC,CAAC;IACnF,CAAC;IACD,GAAG,GAAG,GAAG,CAAC,OAAO,CAAC,WAAW,EAAE,GAAG,MAAM,WAAW,CAAC,CAAC;IAErD,GAAG,CAAC,IAAI,CAAC,mBAAmB,EAAE,GAAG,CAAC,CAAC;IACnC,MAAM,EAAE,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC,kBAAkB,CAAC,CAAC;IACvD,MAAM,GAAG,GAAG,MAAM,GAAG,CAAC,aAAa,CAAC,EAAE,IAAI,EAAE,YAAY,EAAE,CAAC,CAAC;IAC5D,MAAM,SAAS,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;IAC/B,OAAO;QACL,OAAO,EAAE,kBAAkB,UAAU,CAAC,MAAM,oBAAoB,QAAQ,EAAE;QAC1E,OAAO,EAAE,UAAU,CAAC,MAAM;KAC3B,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,kBAAkB,CACtC,QAAgB;IAEhB,MAAM,GAAG,GAAG,MAAM,OAAO,CAAC,QAAQ,CAAC,CAAC;IACpC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,mBAAmB,CAAC,EAAE,CAAC;QACnC,MAAM,IAAI,aAAa,CAAC,cAAc,EAAE,yCAAyC,CAAC,CAAC;IACrF,CAAC;IACD,OAAO,qBAAqB,CAAC,GAAG,CAAC,CAAC;AACpC,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,4BAA4B,CAChD,GAAW;IAEX,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC;IACvC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,mBAAmB,CAAC,EAAE,CAAC;QACnC,MAAM,IAAI,aAAa,CAAC,cAAc,EAAE,yCAAyC,CAAC,CAAC;IACrF,CAAC;IACD,OAAO,qBAAqB,CAAC,GAAG,CAAC,CAAC;AACpC,CAAC"}
|
package/dist/index.d.ts
CHANGED
|
@@ -8,5 +8,5 @@ export { convertDocument } from './convert/index.js';
|
|
|
8
8
|
export { editDocument, editDocx, appendDocxParagraphs, readDocxStructured, updateSpreadsheetCells, readSpreadsheetStructured, } from './edit/index.js';
|
|
9
9
|
export { writeDocument, writeDocx, writePdf, writeSpreadsheet } from './write/index.js';
|
|
10
10
|
export { DocumentError } from './utils/errors.js';
|
|
11
|
-
export type { ParseOptions, ParseOutputMode, ConvertFormat, ConvertOptions, ConvertResult, DocumentInput, DocumentResult, DocumentWarning, DocxStructured, SpreadsheetStructured, LegacyDocStructured, PptxStructured, EditDocumentOp, EditResult, WriteDocumentInput, WriteResult, DocumentErrorCode, } from './types.js';
|
|
11
|
+
export type { ParseOptions, ParseOutputMode, ConvertFormat, ConvertOptions, ConvertResult, DocumentInput, DocumentResult, DocumentWarning, DocxStructured, SpreadsheetStructured, LegacyDocStructured, PptxStructured, ContentBlock, InlineRun, EditDocumentOp, EditResult, WriteDocumentInput, WriteResult, DocumentErrorCode, } from './types.js';
|
|
12
12
|
//# sourceMappingURL=index.d.ts.map
|
package/dist/index.d.ts.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,EAAE,aAAa,EAAE,gBAAgB,EAAE,MAAM,kBAAkB,CAAC;AACnE,OAAO,EAAE,eAAe,EAAE,MAAM,oBAAoB,CAAC;AACrD,OAAO,EACL,YAAY,EACZ,QAAQ,EACR,oBAAoB,EACpB,kBAAkB,EAClB,sBAAsB,EACtB,yBAAyB,GAC1B,MAAM,iBAAiB,CAAC;AACzB,OAAO,EAAE,aAAa,EAAE,SAAS,EAAE,QAAQ,EAAE,gBAAgB,EAAE,MAAM,kBAAkB,CAAC;AACxF,OAAO,EAAE,aAAa,EAAE,MAAM,mBAAmB,CAAC;AAElD,YAAY,EACV,YAAY,EACZ,eAAe,EACf,aAAa,EACb,cAAc,EACd,aAAa,EACb,aAAa,EACb,cAAc,EACd,eAAe,EACf,cAAc,EACd,qBAAqB,EACrB,mBAAmB,EACnB,cAAc,EACd,cAAc,EACd,UAAU,EACV,kBAAkB,EAClB,WAAW,EACX,iBAAiB,GAClB,MAAM,YAAY,CAAC"}
|
|
1
|
+
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,EAAE,aAAa,EAAE,gBAAgB,EAAE,MAAM,kBAAkB,CAAC;AACnE,OAAO,EAAE,eAAe,EAAE,MAAM,oBAAoB,CAAC;AACrD,OAAO,EACL,YAAY,EACZ,QAAQ,EACR,oBAAoB,EACpB,kBAAkB,EAClB,sBAAsB,EACtB,yBAAyB,GAC1B,MAAM,iBAAiB,CAAC;AACzB,OAAO,EAAE,aAAa,EAAE,SAAS,EAAE,QAAQ,EAAE,gBAAgB,EAAE,MAAM,kBAAkB,CAAC;AACxF,OAAO,EAAE,aAAa,EAAE,MAAM,mBAAmB,CAAC;AAElD,YAAY,EACV,YAAY,EACZ,eAAe,EACf,aAAa,EACb,cAAc,EACd,aAAa,EACb,aAAa,EACb,cAAc,EACd,eAAe,EACf,cAAc,EACd,qBAAqB,EACrB,mBAAmB,EACnB,cAAc,EACd,YAAY,EACZ,SAAS,EACT,cAAc,EACd,UAAU,EACV,kBAAkB,EAClB,WAAW,EACX,iBAAiB,GAClB,MAAM,YAAY,CAAC"}
|
|
@@ -0,0 +1,18 @@
|
|
|
1
|
+
import type { ContentBlock, DocxStructured } from '../types.js';
|
|
2
|
+
import type JSZip from 'jszip';
|
|
3
|
+
export type DocxPackageParts = {
|
|
4
|
+
documentXml: string;
|
|
5
|
+
stylesXml?: string;
|
|
6
|
+
numberingXml?: string;
|
|
7
|
+
relsXml?: string;
|
|
8
|
+
footnotesXml?: string;
|
|
9
|
+
endnotesXml?: string;
|
|
10
|
+
};
|
|
11
|
+
/** Parse DOCX package parts into structural blocks. */
|
|
12
|
+
export declare function parseDocxBlocksFromPackage(parts: DocxPackageParts): ContentBlock[];
|
|
13
|
+
/** Parse word/document.xml only (no styles/numbering/links/notes). */
|
|
14
|
+
export declare function parseDocxBlocksFromXml(xml: string): ContentBlock[];
|
|
15
|
+
export declare function docxStructuredFromPackage(parts: DocxPackageParts): DocxStructured;
|
|
16
|
+
export declare function docxStructuredFromXml(xml: string): DocxStructured;
|
|
17
|
+
export declare function docxStructuredFromZip(zip: JSZip): Promise<DocxStructured>;
|
|
18
|
+
//# sourceMappingURL=docx-blocks.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"docx-blocks.d.ts","sourceRoot":"","sources":["../../src/parse/docx-blocks.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,YAAY,EAAE,cAAc,EAAa,MAAM,aAAa,CAAC;AAE3E,OAAO,KAAK,KAAK,MAAM,OAAO,CAAC;AAE/B,MAAM,MAAM,gBAAgB,GAAG;IAC7B,WAAW,EAAE,MAAM,CAAC;IACpB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,WAAW,CAAC,EAAE,MAAM,CAAC;CACtB,CAAC;AAmlBF,uDAAuD;AACvD,wBAAgB,0BAA0B,CAAC,KAAK,EAAE,gBAAgB,GAAG,YAAY,EAAE,CA6DlF;AAED,sEAAsE;AACtE,wBAAgB,sBAAsB,CAAC,GAAG,EAAE,MAAM,GAAG,YAAY,EAAE,CAElE;AAED,wBAAgB,yBAAyB,CAAC,KAAK,EAAE,gBAAgB,GAAG,cAAc,CAOjF;AAED,wBAAgB,qBAAqB,CAAC,GAAG,EAAE,MAAM,GAAG,cAAc,CAEjE;AAED,wBAAsB,qBAAqB,CAAC,GAAG,EAAE,KAAK,GAAG,OAAO,CAAC,cAAc,CAAC,CAkB/E"}
|